Oecologia (2012) 170:867–875
DOI 10.1007/s00442-012-2355-3
GLOBAL CHANGE ECOLOGY - ORIGINAL RESEARCH
High urban population density of birds reflects their timing
of urbanization
Anders Pape Møller • Mario Diaz • Einar Flensted-Jensen •
Tomas Grim • Juan Diego Ibáñez-Álamo • Jukka Jokimäki •
Raivo Mänd • Gábor Markó • Piotr Tryjanowski
Received: 30 December 2011 / Accepted: 26 April 2012 / Published online: 16 May 2012
Ó Springer-Verlag 2012
Abstract Living organisms generally occur at the highest
population density in the most suitable habitat. Therefore,
invasion of and adaptation to novel habitats imply a
gradual increase in population density, from that at or
below what was found in the ancestral habitat to a density
that may reach higher levels in the novel habitat following
adaptation to that habitat. We tested this prediction of
invasion biology by analyzing data on population density
of breeding birds in their ancestral rural habitats and in
matched nearby urban habitats that have been colonized
recently across a continental latitudinal gradient. We estimated population density in the two types of habitats using
Communicated by Esa Lehikoinen.
Electronic supplementary material The online version of this
article (doi:10.1007/s00442-012-2355-3) contains supplementary
material, which is available to authorized users.
A. P. Møller (&)
Laboratoire d’Ecologie, Systématique et Evolution,
CNRS UMR 8079, Université Paris-Sud,
Bâtiment 362, 91405 Orsay Cedex, France
e-mail: [email protected]
M. Diaz
Museo Nacional de Ciencias Naturales (MNCN-CSIC),
28006 Madrid, Spain
extensive point census bird counts, and we obtained
information on the year of urbanization when population
density in urban habitats reached levels higher than that of
the ancestral rural habitat from published records and
estimates by experienced ornithologists. Both the difference in population density between urban and rural habitats
and the year of urbanization were significantly repeatable
when analyzing multiple populations of the same species
across Europe. Population density was on average 30 %
higher in urban than in rural habitats, although density
reached as much as 100-fold higher in urban habitats in
some species. Invasive urban bird species that colonized
urban environments over a long period achieved the largest
increases in population density compared to their ancestral
rural habitats. This was independent of whether species
were anciently or recently urbanized, providing a unique
cross-validation of timing of urban invasions. These results
J. Jokimäki
Arctic Centre, University of Lapland,
FI-96101 Rovaniemi, Finland
R. Mänd
Department of Zoology, Institute of Ecology and Earth Sciences,
University of Tartu, 51014 Tartu, Estonia
E. Flensted-Jensen
Cypresvej 1, 9700 Brønderslev, Denmark
G. Markó
Behavioral Ecology Group, Department of Systematics, Zoology
and Ecology, Eötvös Loránd University, 1/C Pázmány Péter
sétány, Budapest 1117, Hungary
T. Grim
Department of Zoology and Laboratory of Ornithology,
Palacky University, 77146 Olomouc, Czech Republic
G. Markó
Department of Plant Pathology, Corvinus University
of Budapest, Ménesi út 44, Budapest 1118, Hungary
J. D. Ibáñez-Álamo
Departamento de Zoologı́a, Facultad de Ciencias,
Universidad de Granada, Campus Universitario de Fuentenueva
s/n, 18071 Granada, Spain
P. Tryjanowski
Institute of Zoology, Poznań University of Life Sciences,
Wojska Polskiego 71C, 60-625 Poznan, Poland
123
868
suggest that successful invasion of urban habitats was
associated with gradual adaptation to these habitats as
shown by a significant increase in population density in
urban habitats over time.
Keywords Adaptation Birds Cross-validation Invasion Population density
Introduction
Urban habitats cover increasingly large fractions of land,
especially in Europe and North America (European Commission 2006; Schneider et al. 2009), with a further
increase predicted (United Nations 2007). Conversion of
natural habitats into areas partly covered by buildings,
heavily fragmented and with a high level of edges, has
increased dramatically and hence exposed many animals to
human proximity worldwide. Many ecological consequences of habitat conversion to urban areas have long
been recognized, such as altered disturbance regimes, light
conditions, habitat distributions, and species composition
(Rebele 1994; Turner et al. 2004; Alberti 2005; Miller
2005). In addition, urban environments have more
anthropogenic food resources, and the climate of urban
areas differs from that in nearby rural environments
(Gilbert 1989; Rebele 1994) due to the so-called urban heat
island phenomenon (Gilbert 1989). Therefore, urban areas
have longer growing seasons than nearby rural habitats.
Such changes in urban habitats have forced animals and
plants to either adapt or disappear (Coppedge et al. 2001).
Conversion to urban habitats favors the occurrence of
very abundant animal species including birds (Bezzel
1985; Marzluff et al. 2001; Chace and Walsh 2006; Møller
2009). In general, habitat conversion has a negative effect
on rare and specialist species, while it favors generalists
and behaviorally flexible species (Møller 2009). These
changes in abundance and association with humans are
often accompanied by changes in behavior, physiology,
and life history (Dyrcz 1963; Klausnitzer 1989; Møller
2008b, 2009; Møller et al. 2009; Zhang et al. 2011). This
phenomenon of organismal changes we term the biological
process of urbanization. Urban birds may be better adapted
to urban environments than rural conspecifics, even when
these urban birds are introduced to a novel area, as was the
case when blackbirds Turdus merula were introduced to a
city that previously did not have this species breeding
(Graczyk 1974).
Some species are better able to live in and colonize
urban habitats than others (Klausnitzer 1989; Stephan
1999; Gliwicz et al. 1994; Anderies et al. 2007; Møller
2009, 2010a; Carrete and Tella 2011). Because environmental conditions have changed due to habitat conversion,
123
Oecologia (2012) 170:867–875
species that have lived for a long time in urban environments such as feral pigeons Columba livia domestica,
house sparrows Passer domesticus, and blackbirds have
become adapted to urban environments, causing such urban
populations to reach higher population densities than
populations in nearby rural habitats. For example, blackbirds now reach densities that are more than two orders of
magnitude higher in urban habitats compared to the
ancestral forested rural habitats, where this species exclusively bred until 180 years ago (Dyrcz 1963; Klausnitzer
1989; Luniak et al. 1990; Stephan 1999; Gliwicz et al.
1994).
Species of birds that currently live in cities can be categorized as anciently associated with humans for thousands
of years, such as feral pigeons and house sparrows, because
they are pre-adapted to urban habitats with many and large
buildings (e.g., Parmelee 1959 on the birds of the bible;
Gesner 1669 on birds in general; Møller 1992 on pre-fossil
finds of barn swallows Hirundo rustica from a Danish flint
mine), or they have more recently become established in
cities from rural open parkland habitats. Thus urbanization
of birds may either constitute a gradual process that have
been going on for a different period of time for different
species, or it may differ qualitatively between species that
have lived for millennia in urban environments.
Different criteria have been used for successful invasion
of urban environments. Møller (2009) used the criterion for
urbanization of breeding birds that at least one urban
population has higher population density than nearby rural
populations. Croci et al. (2008) defined urbanized species
of those that were able to breed in city centers. Finally,
Evans et al. (2010) analyzed British census data and used
the population density in urban habitats and the difference
in density between all urban and all rural habitats to
investigate determinants of degree of urbanization. Here,
we extend this approach by comparing nearby urban and
rural habitats in a paired design that controls statistically
for the fact that urban and rural habitats may differ in many
other respects than urbanization. Indeed, the approach that
we adopt here relies on the fact that nearby areas differing
in whether they are urban or rural will be similar in other
respects such as habitat, geology, weather, and history of
human exploitation including pollution. However, it
remains to be tested whether adaptation to the invaded
novel habitat as reflected by population density generally
increases over time, although such a test would provide
much needed cross-validation of a central assumption in
invasion biology (e.g., Davis 2009; Lockwood et al. 2006).
Such an increase may not necessarily be linear, but rather
be gradual followed by a rapid increase (e.g., Stephan
1999; Vuorisalo et al. 2003).
The objectives of this study were to test for gradual
adaptation to novel habitats in invasive species by
Oecologia (2012) 170:867–875
analyzing extensive data on breeding birds occupying
urban environments. To this end, we first provide two
independent estimates of the degree of urbanization of
different species: the difference in population density
between urban and nearby rural habitats, and the estimated
year of urbanization. The difference in population density
between urban and nearby rural habitats provides an estimate of the extent to which a species has adapted to the
urban environment, with a value of zero indicating no
difference, a positive value an increase in density in urban
habitats, and a negative value that the species has not
become adapted to urban habitats. Second, we assess to
which extent these measures of degree of urbanization are
repeatable among cities, thereby testing if one estimate
from a comparison between an urban and a nearby rural
area has any predictive power for estimating the degree of
urbanization by the same species in other cities elsewhere
in Europe. Third, we test if urban species of birds can be
considered to consist of two different categories of species
that have either been urbanized for millennia or only more
recently. Finally, we assess whether there is gradual
adjustment to urban environments by testing if the difference in population density between urban and rural habitats
increases with the time since known time of initial urbanization (based on literature information and data collected
by local ornithologists). Here, we assess these predictions,
using birds in nine pairs of European cities and nearby rural
habitats as a model system for invasion biology. While
urban ecosystems may differ from other ecosystems by
their high and unpredictable disturbance regimes, they are
not so different from high and unpredictable disturbance in
agricultural, forest, or lagoon ecosystems exploited by
humans.
869
Fig. 1 Location of the nine paired localities with urban and rural
study sites for urbanization of birds
simple operational definition was also adopted in other
studies (e.g., Klausnitzer 1989; Gliwicz et al. 1994;
Stephan 1999), and our definitions of urban (percent of
built-up area [50, building density [10/ha, and residential
human density [10/ha) and rural habitats (percent of builtup area 5–20, building density [2.5/ha, and residential
human density 1–10/ha) follow the suggestion made by
Marzluff et al. (2001). A list of cities, coordinates, and their
human population sizes is provided in electronic appendix
Table 1.
Materials and methods
Ancient or recently urbanized species
Study areas
We recorded population density of breeding birds in nine
cities (each paired with a nearby rural area) across Europe
by using consistent methods across all spatial replicates
(Fig. 1). The distance between urban and rural study sites
was 1–20 km, which can be considered small, given the
flying ability of birds. The benefits of this matched design
is that neighboring rural and urban study sites will share
most characteristics including weather, altitude, soil, and
many others.
All urban study sites included areas with multi-storey
buildings, single-family houses, roads, and parks, while
nearby rural areas had open farmland and woodland and
did not contain continuous urban elements like multi-storey
buildings, single-family houses, roads, and parks. This
We classified all species as being ancient urban species if
they are known to have inhabited urban settlements since
ancient times in the Mediterranean basin and the Middle
East. The species belonging to this category were rock
dove, jackdaw Corvus monedula, barn swallow, house
martin Delichon urbicum, house sparrow, spotless starling
Sturnus unicolor, and starling S. vulgaris. All other species
were classified as recently urbanized.
Estimating urbanization level of species I:
timing of urbanization
Bird species that have been urbanized for a long period will
have spent more generations in urban areas, and hence
will have had a longer time to adapt to this environment.
123
870
We estimated the approximate year of urbanization in the
different cities, as described in detail by Møller (2008a,
2009, 2010a). Timing of urbanization will result from
colonization followed by establishment or extinction and
re-colonization. Obviously, information is rarely available
for such processes, and empirical information on development of urban population sizes is also scarce (but see
Stephan 1999; Vuorisalo et al. 2003). In the following, we
assume that colonization of urban environments can be
approximated from observations by keen ornithologists that
habitually closely follow changes in composition and distribution of birds. Any heterogeneity in colonization processes or increase in population size will cause noise in the
data, and ultimately make it more difficult to discern any
patterns. We estimated the year when different species
became urbanized using two different approaches. First, we
asked keen amateur ornithologists living in our study areas
to state which year different species of birds were first
recorded breeding in urban areas. This estimate was used as
an approximate year of urbanization, with a conservative
value of 1950 assigned to species that were known to breed
in urban habitats before the observers started watching
birds. Previously, Møller (2008a) asked two keen Danish
amateur ornithologists (William C. Årestrup and E. Flensted-Jensen) living in a study area in Denmark to state
when different species of birds were first recorded breeding
in urban areas. An approximate year of urbanization was
recorded, with a value of 1950 assigned to species that
were known to breed in urban habitats before the two
observers started watching birds. Independently, Møller
also recorded these years for all species before asking for
estimates from the two independent observers. These three
datasets were highly consistent in assignment of year (all
three Pearson r [ 0.96, n = 44 species). Second, we
recorded timing of invasion of urban environments from
old published records. Although urbanization is likely to
have occurred much earlier for many species such as house
sparrow and rock pigeon, these estimates are conservative.
Previouslyn Møller (2008a) recorded the timing of invasion
of urban environments in Copenhagen, Denmark, and
Paris, France, from old published records (Gram 1908;
Fløystrup 1920, 1925 for Copenhagen, and an extensive
literature survey of ornithological journals and handbooks
such as Cramp and Perrins 1977–1994 and Glutz and Bauer
1985–1997). If the year of urbanization was before records
reported in these sources, Møller assigned 1850 as the year
of urbanization (because Gram’s observations date back to
1850, and many species were urbanized before the start of
his study). Although urbanization is likely to have occurred
much earlier for many species such as house sparrow and
rock pigeon, these estimates are conservative. This dataset
provided independent estimates of the time of urbanization,
but these were still strongly positively correlated with
123
Oecologia (2012) 170:867–875
Møller’s own estimates from Northern Jutland (Pearson
r = 0.72, n = 41 species). For all populations with
urbanized populations, we used an estimate of 2010 as the
year of urbanization for populations that were not yet
urbanized. We included in the analyses all populations for
which we could obtain information on year of urbanization.
Estimating urbanization level of species II: density
in urban and rural areas
Bird species that have higher densities in urban versus
rural areas could be considered to have adapted successfully because a hallmark of adaptation to local conditions
is high local and global population densities (Brown 1995;
Brown and Lomolino 1998). We performed standard point
count censuses of breeding birds with unlimited recording
distance (e.g. Vořı́šek et al. 2010), twice with an interval
of 3–4 weeks, during spring 2010 in both urban and rural
habitats in all study locations. Point counts provide highly
reliable estimates of relative population density that is
comparable among habitats (Vořı́šek et al. 2010). First,
we placed 25–50 points (depending on the size of the
particular urban area because some smaller urban areas
did not allow allocation of 50 points) in each urban and
rural study plot at a distance of 100 m between consecutive points, using a stratified random sampling design to
ensure that all locally available habitats were included.
The exact location of each point was determined with a
GPS, allowing us to make the second census in exactly the
same sites as the first census. Second, we made a first
census in early spring starting early April in southern
Spain, delaying the census at higher latitudes so it was
completed in northern Finland in late May. The census
started at local sunrise, while remaining 5 min at each
point recording all birds seen or heard. Censuses were
finished before 1000 hours (Vořı́šek et al. 2010). Censuses started on alternate days in urban and rural study
plots ensuring that there was no difference in timing of
censuses between habitats.
Vegetation cover (trees, shrubs, herbs, and grass) and
cover with buildings and other man-made structures were
quantified to the nearest 10 % in the field within 50 m of
each survey point. In estimates of population density, we
obtained quantitatively similar results when controlling or
not controlling for differences in coverage for the three
vegetation layers (results not shown).
As an estimate of relative population density in urban
compared to rural habitats, we used log10-transformed
population density in urban areas minus log10-transformed
population density in rural areas, adding a constant of 0.01
to avoid problems with a few estimates of zero. The density
for each of the two habitats was the mean number of
individuals recorded per census point during the two point
Oecologia (2012) 170:867–875
count censuses made in 2010. The data are provided in
electronic appendix Table 2.
Statistical analyses
All analyses were made with JMP (SAS Institute 2000). We
log10-transformed population densities from point counts,
adding a constant of 0.01 in order to be able to include population densities of zero to achieve distributions that did not
differ from normality. The variable classifying species as
anciently or recently urbanized was treated as a dichotomous
dummy variable with levels of 0 (recent) or 1 (ancient).
We tested for repeatability of difference in population
density and year of urbanization using species as a predictor
for those species that had at least two pairs of urban and rural
populations, using statistics in Falconer and Mackay (1996).
We tested if the difference in population density
between urban and nearby rural habitats could be predicted
from the year of urbanization, using a GLM with difference
in density as the response variable and year of urbanization
and species as a random effect to avoid problems of nonindependence of different cases of urbanization by the
same species. This might not constitute a serious problem
because Evans et al. (2009) have shown for the blackbird
that different urban populations have become established
independently rather than through migration from a single
urban source population to other urban areas, and several
studies have shown that even neighboring populations of
urban birds are genetically differentiated (Baratti et al.
2009; Björklund et al. 2010; Fulgione et al. 2000;
Rutkowski et al. 2005), suggesting recent divergence even
within urban areas. Finally, we included the dichotomous
dummy variable reflecting ancient or recent urbanization in
these models. We entered the interaction between this
dummy variable and year of urbanization, but also constructed a model that excluded anciently urbanized species
to test for a qualitative difference between anciently and
recently urbanized species.
Values reported are means (SE).
Results
Estimates of degree and timing of urbanization
The mean difference in density between urban and rural
habitats on a log-scale was 0.112 (SE = 0.039, range -1.656
to ?2.021, n = 250 populations, differing significantly from
zero (one-sample t test, t = 2.90, df = 249, P = 0.004). This
implies that population density in urban areas varied from
2.2 % of the density of the same species in rural habitats to
105 times the density in rural habitats. The mean of 0.112
implies that density in urban habitats was 1.29 times the
871
density in rural habitats. The five species with the relatively
largest density in urban habitats were tree sparrow Passer
montanus in Denmark (105 times density in rural habitat),
house sparrow in Norway (83 times), jackdaw Corvus
monedula in Poland (77 times), rock dove in Finland (63
times) and black-headed gull Larus ridibundus in Finland (36
times). The frequency distribution of differences in population density only deviated marginally from a normal distribution (Shapiro–Wilk test, W = 0.971, P = 0.012).
Breeding population density in urban habitats was positively
correlated with density in nearby rural habitats [F = 115.71,
df = 1, 248, r2 = 0.32, P \ 0.0001, slope (SE) = 0.766
(0.071)]. This implies that urban species already had high
population densities in their ancestral rural habitats, under the
assumption that current density in rural habitats also reflects
past density in these habitats. Difference in density was not
significantly related to latitude in a model that also included
species (partial effect of latitude: F = 0.14, df = 1, 189,
P = 0.70).
The first year of urbanization ranged from 1850 to 2010,
mean (SE) = 1966 (4 years), n = 174. If we excluded all
values from 2010 (implying that the species had not yet
become urbanized), the mean year was 1938 (4 years), range
1850 to 2005. There was a weak negative relationship between
year of urbanization and latitude in a model that also included
species [partial effect of latitude: F = 4.77, df = 1, 122,
r2 = 0.03, P \ 0.0001, slope (SE) = -0.960 (0.439)].
Repeatability of degree and timing of urbanization
The difference in density was significantly repeatable among
populations for the species with at least 2 estimates (F = 2.44,
df = 59, 190, r2 = 0.43, P \ 0.0001) with a repeatability of
0.26 (SE = 0.05). The estimates of repeatability were not
significantly different for anciently and recently urbanized
species [ancient: F = 2.02, df = 7, 32, r2 = 0.31, P = 0.08,
R = 0.17 (SE = 0.10); recent: F = 2.47, df = 51, 158,
r2 = 0.44, P \ 0.0001, R = 0.27 (SE = 0.05)].
Year of urbanization was significantly repeatable among
populations for the species with at least two estimates
(F = 3.66, df = 50, 123, r2 = 0.60, P \ 0.0001) with a
repeatability estimate of 0.44 (SE = 0.06). The estimates of
repeatability were not significantly different for anciently
and recently urbanized species [ancient: F = 3.80, df = 7,
20, r2 = 0.57, P = 0.009, R = 0.44 (SE = 0.15); recent:
F = 1.82, df = 51, 102, r2 = 0.47, P = 0.005, R = 0.21
(SE = 0.07)].
Relationship between the difference in density and year
of urbanization
The difference in population density between urban and
rural habitats was significantly negatively related to year of
123
872
urbanization [Fig. 2; F = 52.09, df = 1,181, r2 = 0.22,
P \ 0.0001, slope (SE) = -0.0056 (0.0008)]. This result
implies that a 100-year difference in timing of urbanization
was associated with a factor 3.64 difference in population
density between urban and rural habitats (SE = 1.20).
There was no significant additional partial effect of latitude
(F = 0.58, df = 1,180, P = 0.45). If we excluded all
populations with an estimated year of urbanization of 2010,
there was still a highly significant effect of year of urbanization [F = 18.83, df = 42,63, r2 = 0.23, P \ 0.0001,
slope (SE) = -0.0055 (0.0013)].
If we entered species as a random effect in this model, to
account for the possible non-independence of multiple
populations of the same species, the model explained 36 %
of the variance (F = 1.69, df = 60, 122, r2 = 0.36,
P = 0.0073). The partial effect of year of urbanization was
highly significant [F = 39.74, df = 1, 122, r2 = 0.25,
P \ 0.0001, slope (SE) = -0.0052 (0.0008)]. The slope of
-0.0052 implies that a 100-year difference in timing of
urbanization was associated with a difference in density by
a factor 3.31 (SE = 1.20). There was no significant additional partial effect of latitude (F = 0.14, df = 1,189,
P = 0.70).
If we related difference in density to year urbanized,
population and the variable reflecting ancient or recent
urbanization, we find a model that explains 29 % of the
variance (F = 7.17, df = 10,172, r2 = 0.29, P \ 0.0001).
The largest effect was for year urbanized [F = 33.97,
df = 1,172, P \ 0.0001, slope (SE) = -0.0055 (0.0009)],
followed by a marginal effect of population (F = 1.93,
df = 8,172, P = 0.06) and a small, non-significant effect
of ancient or recent urbanization (F = 1.12, df = 1,172,
P = 0.29). Likewise, the interaction between year of
urbanization and ancient or recent urbanization was small
and non-significant (F = 1.15, df = 1,171, P = 0.28).
Finally, a model that excluded all anciently urbanized
Fig. 2 Difference in breeding population density between urban and
rural habitats in relation to estimated year of urbanization in different
populations of species of birds. The difference in density is
log10(population density in urban habitats) - log10(population density in rural habitats)
123
Oecologia (2012) 170:867–875
species explained 54 % of the variance (F = 2.33,
df = 52,102, r2 = 0.54, P \ 0.0001), with significant
effects of year urbanized [F = 10.09, df = 1,102, P =
0.002, slope (SE) = -0.0038 (0.0012)] and species
(F = 1.85, df = 1,102, P = 0.005). These analyses suggest that anciently urbanized species do not differ qualitatively from recently urbanized species, and that overall
species that have been urbanized over a long period have a
greater difference in population density between urban and
rural habitats.
Discussion
We tested the assumptions of invasion biology that species
that had invaded a novel environment had achieved higher
population density in the invaded environment, and that the
increase in density reflected time since invasion of the
novel habitat. We did so by analyzing extensive data on
density and timing of invasion of urban habitats by birds.
The mean difference in population density of the same
species between urban and nearby rural habitats was
positive, implying that most species had higher densities in
urban habitats. The difference was significantly repeatable,
albeit with a small repeatability, suggesting that, when
population density differed between urban and rural habitats to a certain degree in one city in Europe, this difference
tended to be of similar magnitude in other cities. Likewise,
the estimated year of urbanization was significantly
repeatable among populations, implying that, when a species became urbanized in one city in a given year, other
cities tended to become urbanized at a similar time elsewhere in Europe. Finally, we showed that the difference in
population density between urban and nearby rural habitat
was significantly negatively related to the estimated year of
urbanization, suggesting that adaptation to urban environments results in a gradual increase in population density in
the invaded environment over time, and that the difference
in density is a reasonably good proxy for time of urbanization. Importantly, this relationship between difference in
population density and year of urbanization was qualitatively similar in anciently and recently urbanized species,
suggesting that anciently urbanized species that have been
associated with humans for millennia show qualitatively
similar patterns of urbanization as recently urbanized species. This provides empirical tests of critical assumptions
of invasion biology, but also of repeatability of invasiveness across large geographical scales, something that has
rarely if ever been tested (e.g., Davis 2009; Lockwood
et al. 2006).
Urban birds have large population densities relative to
the density of the same species in nearby rural habitats
(e.g., Dyrcz 1963; Klausnitzer 1989; Luniak et al. 1990;
Oecologia (2012) 170:867–875
Stephan 1999; Gliwicz et al. 1994; Evans et al. 2010). On
average, across the species and populations sampled here,
we found an increase in population density in urban habitats by almost 30 % compared to nearby rural habitats. In
addition, population density in novel urban habitats was
significantly positively correlated with density in ancestral
rural habitats. This effect implies that bird species that have
successfully invaded urban habitats already had high population densities in their ancestral rural habitats. Indeed, a
pairwise comparative analysis of phylogenetically independent cases of urbanization by birds in the Western
Palearctic showed that range size, population size and
population density were consistently higher in the species
that became urbanized compared to a non-urbanized closely related species (Møller 2009). The difference in population density between urban and rural habitats reached
more than two orders of magnitude, and similarly large
differences have been reported elsewhere (Dyrcz 1963;
Klausnitzer 1989; Luniak et al. 1990; Stephan 1999;
Gliwicz et al. 1994; Evans et al. 2010). These differences
in population density could also be generalized across
populations of the same species, as shown by a statistically
significant repeatability. Repeatability estimates in genetics
are use to set an upper limit to the heritability (Falconer
and Mackay 1996). However, repeatabilities can also be
estimated in other contexts such as ecological studies. With
values ranging from 0 to 1, a value of 0.26 is modest,
although ecological studies consistently have small to
intermediate effect sizes hence implying low repeatabilities
(Møller and Jennions 2002). These findings also relate to
theoretical models of population density in urban environments with low predation pressure and more predictable
and abundant food resources (Sochat 2004; Anderies et al.
2007). In competition models with two phenotypes that
differ in foraging ability, short periods between pulses of
resources and low predation rates result in weaker competitors, while longer periods and high predation rates
favor stronger competitors (Sochat 2004; Anderies et al.
2007). Indeed, species differing in susceptibility to predation are differentially successful in urbanization (Møller
2009), and urban birds with higher population densities
than rural populations of the same species differ in terms of
song post position, the fraction of nests found indoors
inside buildings, and escape behavior by birds captured by
humans (Møller 2008a, 2010a, b, 2011; Møller and IbáñezÁlamo 2012). Interestingly, these differences in antipredator behavior are strongly correlated with time since
urbanization. The findings that we report here suggest that
time since urbanization may be an additional factor to
consider in such models. Point counts provide reliable
information about relative density, especially when comparing populations of the same species in different habitats
(Vořı́šek et al. 2010). However, the precision of population
873
densities based on 25–50 census points is limited, suggesting that the repeatability of difference in population
density constitutes a conservative estimate. The fact that
population density in urban habitats could be predicted by
population density in rural habitats in a model that
explained 32 % of the variance implies that the two indices
of urbanization (urban density and difference in density
between urban and rural habitats) provide qualitatively
similar information.
Urbanization has been ongoing for a long time. For
example, many species of insects, birds, and mammals
have co-habited with humans for millennia (Klausnitzer
1989). Anciently urbanized birds such as rock doves, barn
swallows, house martins, house sparrows, and jackdaws are
known to have been associated with human settlements in
the Middle East for more than 5,000 years (e.g., Parmelee
1959; Summers-Smith 1963). It is known that blackbirds
were already very common in the city of Rome in the
1820s (Bonaparte 1828). Likewise, many birds have bred
indoors inside human habitation for millennia, and species
that are adapted to human proximity are also those that
have been urbanized for a long time (Møller 2010b). Here,
we analyzed the year of urbanization as estimated by
amateur ornithologists and published reports in the ornithological literature. We found significant repeatability in
year of urbanization among populations, with an estimate
of 0.44. This implies that more than 40 % of the variance in
year of urbanization is accounted for by species. Thus, the
same species tended to become urbanized at the same time
in different parts of Europe. Importantly, we found no
significant difference between anciently and recently
urbanized species, suggesting that these two categories of
species behaved qualitatively similar to urbanization. This
is an important finding suggesting that the process of
association between birds and humans is a continuous
process dating back thousands of years.
There is a currently increasing interest in the study of
biological aspects of urbanization (e.g., Grimm et al. 2008),
but also of invasion biology (e.g., Davis 2009; Lockwood
et al. 2006). Here, we have provided significant evidence
for adaptation to novel environments by invasive urban
bird species, because species that invaded urban environments a long time ago are also the species that have
achieved the largest increases in population density compared to their ancestral rural habitats. This finding is ecologically important because species with high local
population densities tend to have high global population
sizes (Brown 1995; Brown and Lomolino 1998). There
have been few attempts to explicitly test to which extent
different metrics of urbanization provide similar information. Møller (2008a, 2010a) has shown that the estimated
year of urbanization of different species of birds is highly
consistent when using observations by different observers,
123
874
and also when comparing observations and published
records of date of urbanization. Here, we have taken this
cross-validation approach one step further by relating the
difference in population density among all populations to
the year of urbanization, after controlling for the fact that
different populations of the same species may not constitute statistically independent observations unless urbanization has occurred independently in different cities (Evans
et al. 2009). The correlation between the two metrics of
urbanization of bird populations accounted for 22–25% of
the variance in the data. This provides evidence for consistency in different estimates of urbanization. Clearly
22–25 % of the variance is not a large amount, although
again this should be viewed in the light of the inherent
uncertainty of the two estimates. We can judge the relative
magnitude of these effects by comparing the estimates with
the amount of variance explained by all meta-analyses in
the biological sciences that was in the order of 5–10 % of
the variance (Møller and Jennions 2002).
The implications of this study are that urbanized birds
have shown gradual adjustment to urban environments, as
shown by relative larger population densities in urban
habitats in species that have been urbanized for a long time.
In addition, we can predict a small, but significant, amount
of variance in difference in population density and year of
urbanization from just a single estimate according to the
estimates of repeatability. Obviously, the precision of such
estimates will improve with the number of populations
used. The findings also imply that we can use differences in
population density and year of urbanization for further
studies, knowing that these estimates provide reliable
information on two aspects of urbanization. Finally, several
studies have shown significant genetic differentiation
between populations in rural and urban habitats or among
populations inhabiting different urban areas (e.g., Baratti
et al. 2009; Björklund et al. 2010; Evans et al. 2009;
Fulgione et al. 2000; Rutkowski et al. 2005). Because
Crispo and Hendry (2005) have shown that isolation by
distance increases with time since colonization, we make
the additional prediction that the degree of such differentiation will be greater in species that show large differences
in population density between urban and rural habitats and
for populations that have been urbanized for a long time.
However, anciently and recently urbanized species should
only differ in degree rather than kind, as shown by the lack
of effect of ancient or recent urbanization on the relationship between year of urbanization and difference in density
between urban and nearby rural habitats in the present
work.
In conclusion, urban birds provide an ideal model system for the study of invasion of novel environments,
showing consistency in timing of invasion and extent of
adaptation to these novel environments. In addition, we
123
Oecologia (2012) 170:867–875
have shown that both the difference in population density
between urban and rural habitats and the estimated year of
urbanization are repeatable among cities in Europe, and
that different species on average have almost 30 % higher
population density in urban habitats. Furthermore, the
difference in population density between urban and rural
habitats is negatively correlated with the estimated year of
urbanization, independent of whether species are anciently
or recently urbanized. This result provides cross-validation
of the assumption that different measures of the degree of
urbanization provide similar and reliable information.
These findings have implications for future studies of
urbanization, and for invasion biology in general, because
the high and unpredictable level of disturbance in urban
ecosystems makes urban ecosystems similar to other ecosystems exploited by humans such as farmland, forests, and
lagoons.
Acknowledgments Two anonymous reviewers provided helpful
suggestions and proposed the analyses of effects of ancient or recent
urbanization. R.M. was financially supported by the Estonian Ministry of Education and Science (target-financing project number
0180004s09) and the European Union through the European Regional
Development Fund (Center of Excellence FIBIR). T.G. was supported
by the Human Frontier Science Program (RGY69/07) and
MSM6198959212. J.J. received support from the EU Regional
Development Fund via the project ‘‘Rovaniemen kaupunkilintualtas’’.
G.M. was supported by TÁMOP-4.2.1./B-09/1-KMR-2010-0005
grant. E. Leibak, M. Martı́n-Vivaldi, A. Tinaut, and J. M. Pleguezuelos kindly provided information on timing of urbanization. I. Aus, Z.
Bajor, A. Dvorska, and A. Jair helped with fieldwork.
References
Alberti M (2005) The effects of urban patterns on ecosystem function.
Int Reg Sci Rev 28:168–192
Anderies JM, Katti M, Shochat E (2007) Living in the city: resource
availability, predation, and bird population dynamics in urban
areas. J Theor Biol 247:36–49
Baratti M, Cordaro M, Dessi-Fulgheri F, Vannini M, Fratini S (2009)
Molecular and ecological characterization of urban populations
of the mallard (Anas platyrhynchos) in Italy. Italy J Zool 76:330–
339
Bezzel E (1985) Birdlife in intensively used rural and urban
environments. Ornis Fenn 62:90–95
Björklund M, Ruiz I, Senar JC (2010) Genetic differentiation in the
urban habitat: the great tits (Parus major) of the parks of
Barcelona city. Biol J Linn Soc 99:9–19
Bonaparte C-L (1828) Ornithologie compare de Rome et de
Philadelphie. Rome
Brown JH (1995) Macroecology. Chicago University Press, Chicago
Brown JH, Lomolino MV (1998) Biogeography, 2nd edn. Sinauer,
Sunderland
Carrete M, Tella JL (2011) Inter-individual variability in fear of
humans and relative brain size of the species are related to
contemporary urban invasion in birds. PLoS ONE 6(4):e18859
Chace JF, Walsh JJ (2006) Urban effects on native avifauna: a review.
Landsc Urban Plan 74:46–69
Oecologia (2012) 170:867–875
Coppedge BR, Engle DM, Fuhlendorf SD, Masters RE, Gregory MS
(2001) Urban sprawl and juniper encroachment effects on
abundance of wintering passerines in Oklahoma. In: Marzluff
JM, Bowman R, Donelly R (eds) Avian ecology and conservation in an urbanizing world. Kluwer, Dordrecht, pp 225–242
Cramp S, Perrins CM (eds) (1977–1994) The birds of the Western
Palearctic, vols 1–9. Oxford University Press, Oxford
Crispo E, Hendry AP (2005) Does time since colonization influence
isolation by distance? A meta-analysis. Conserv Genet
6:665–682
Croci S, Butet A, Clergeau P (2008) Does urbanization filter birds on
the basis of their biological traits? Condor 110:223–240
Davis MA (2009) Invasion biology. Oxford University Press, Oxford
Dyrcz A (1963) Comparative studies on the avifauna of wood and
park. Acta Ornithol 12:177–208
European Commission (2006) Urban sprawl in Europe. European
Environmental Agency, Copenhagen
Evans KL, Gaston KJ, Frantz AC, Simeoni M, Sharp SP, McGowan
A, Dawson DA, Walasz K, Partecke J, Burke T, Hatchwell BJ
(2009) Independent colonization of multiple urban centres by a
formerly forest specialist bird species. Proc R Soc Lond B
276:2403–2410
Evans KL, Chamberlain DE, Hatchwell BJ, Gregory RD, Gaston KJ
(2010) What makes an urban bird? Glob Change Biol 17:32–44
Falconer DS, Mackay TFC (1996) Introduction to quantitative
genetics, 4th edn. Longman, New York
Fløystrup A (1920) Fugleliv i Kjøbenhavn: Iagttagelser fra Østre
Anlæg og Botanisk Have. Dansk Orn Foren Tidsskr 14:1–60
Fløystrup A (1925) Fugleliv i Kjøbenhavn: Fortsatte iagttagelser fra
Østre Anlæg og Botanisk Have. Dansk Orn Foren Tidsskr
19:1–18
Fulgione D, Rippa D, Procaccini G, Milone M (2000) Urbanisation
and the genetic structure of Passer italiae (Viellot 1817)
populations in the south of Italy. Ethol Ecol Evol 12:123–130
Gesner C (1669) Vollkommenes Vogelbuch. Schlütersche, Hannover
(reprint 1981)
Gilbert OL (1989) The ecology of urban habitats. Chapman and Hall,
London
Gliwicz J, Goszczynski J, Luniak M (1994) Characteristic features of
animal populations under synurbanization: the case of the
blackbirds and the striped field mouse. Mem Zool 49:237–244
Glutz von Blotzheim UN, Bauer KM (eds) (1985–1997) Handbuch
der Vögel Mitteleuropas, vols 1–14. AULA, Wiesbaden
Graczyk R (1974) The experiment of settling of urbanized population
of Poznan blackbird (Turdus merula L.) to Kiev (USSR) and
examination of certain elements of innate behaviour. Roczn
Akad Roln Pozn 65:49–64
Gram C (1908) Fuglelivet i København og omegn for halvhundrede
aar siden. Dansk Orn Foren Tidsskr 3:27–36
Grimm NB, Faeth SH, Golubiewski NE, Redman CL, Wu J, Bai X,
Briggs JM (2008) Global change and the ecology of cities.
Science 319:756–760
Klausnitzer B (1989) Verstädterung von Tieren. Neue BrehmBücherei, Wittenberg Lutherstadt
Lockwood J, Hoopes M, Marchetti M (2006) Invasion ecology.
Blackwell, Oxford
Luniak M, Mulsow R, Walasz K (1990) Urbanization of the European
blackbird: expansion and adaptations of urban population. In:
Luniak M (ed) Urban ecological studies in Central and Eastern
Europe. Ossolineum, Warsaw, pp 87–199
875
Marzluff JM, Bowman R, Donnelly R (2001) A historical perspective
on urban bird research: trend, terms, and approaches. In:
Bowman R, Donelly R, Marzluff JM (eds) Avian Ecology and
Conservation in an Urbanizing World. Kluwer, Norwell,
pp 20–47
Miller JR (2005) Biodiversity conservation and the extinction of
experience. Trends Ecol Evol 20:430–434
Møller AP (1992) The hirundines. Natur og Museum 31:1–32
Møller AP (2008a) Flight distance of urban birds, predation and
selection for urban life. Behav Ecol Sociobiol 63:63–75
Møller AP (2008b) Flight distance and population trends in European
breeding birds. Behav Ecol 19:1095–1102
Møller AP (2009) Successful city dwellers: a comparative study of
the ecological characteristics of urban birds in the Western
Palearctic. Oecologia 159:849–858
Møller AP (2010a) Interspecific variation in fear responses predicts
urbanization in birds. Behav Ecol 21:365–371
Møller AP (2010b) The fitness benefit of association with humans:
elevated success of birds breeding indoors. Behav Ecol 21:913–
918
Møller AP (2011) Song post height in relation to predator diversity
and urbanization. Ethology 117:529–538
Møller AP, Ibáñez-Álamo JD (2012) Capture behavior of urban birds
provides evidence of predation being involved in urbanization.
Anim Behav (in press)
Møller AP, Jennions MD (2002) How much variance can be explained by
ecologists and evolutionary biologists? Oecologia 132:492–500
Møller AP, Erritzøe J, Karadas F (2009) Levels of antioxidants in
rural and urban birds and their consequences. Oecologia 163:35–
45
Parmelee A (1959) All the birds of the bible: their stories,
identification and meaning. Lutterworth, London
Rebele F (1994) Urban ecology and special features of urban
ecosystems. Glob Ecol Biogeogr Lett 4:173–187
Rutkowski R, Rejt L, Gryczynska-Siematkowska A, Jagolkowska P
(2005) Urbanization gradient and genetic variability of birds:
example of kestrels in Warsaw. Berkut 14:130–136
SAS Institute Inc (2000) JMP. SAS Institute, Cary
Schneider A, Friedl MA, Potere D (2009) A new map of global urban
extent from MODIS satellite data. Environ Res Lett. doi:
10.1088/1748-9326/4/4/044003
Sochat E (2004) Credit or debit? Resource input changes population
dynamics of city slicker birds. Oikos 106:622–626
Stephan B (1999) Die Amsel. Neue Brehm-Bücherei, WittenbergLutherstadt
Summers-Smith D (1963) The house sparrow. Collins, London
Turner WR, Nakamura T, Dinetti M (2004) Global urbanization and
the separation of humans from nature. Bioscience 54:585–590
United Nations (2007) World urbanization prospects: the 2007
revisions. United Nations, New York
Vořı́šek P, Klvanova A, Wotton S, Gregory RD (2010) A best
practice guide for wild bird monitoring schemes. European
Union, Bruxelles
Vuorisalo T, Andersson H, Hugg T, Lahtinen R, Laaksonen H,
Lehikoinen E (2003) Urban development from an avian
perspective: causes of hooded crow (Corvus corone cornix)
urbanisation in two Finnish cities. Landsc Urban Plan 62:69–87
Zhang S, Lei F, Liu S, Li D, Chen C, Wang P (2011) Variation in
baseline corticosterone levels of tree sparrow (Passer montanus)
populations along an urban gradient. J Ornithol 152:801–806
123